Generative artificial intelligence-powered enterprise automation startup Orby AI Inc. today revealed it has closed on a $30 million early-stage funding round, fueling its mission to eliminate mundane processes from work.

Today’s Series A round was led by New Enterprise Associates, Wing VC and Jeffrey Katzenberg’s WndrCo. Pear VC also participated, having co-led the company’s original $4.5 million seed funding round in May 2023.

Orby AI wants to use its advanced generative AI models to tackle some of the headaches in enterprise automation, in particular the problems associated with automating more complex tasks.

The startup explains that most businesses today have to perform various tasks, such as data entry, document processing and validation of complex forms, contracts, invoices and so on. Though existing automation tools can do these basic tasks, they aren’t always successful, as they lack the necessary sophistication to be able to complete more complex workflows. That’s because these legacy tools are only designed to perform rules-based automation, which is a system that applies human-made rules to store, sort and manipulate data.

One of the problems with rules-based automation is that it tends to be rigid, unable to handle technical or formatting changes. For instance, if a rules-based automation tool is tasked with transferring data from one interface to another, but there are differences in how that data is formatted, it probably won’t be able to complete that task, the startup explained.

Orby AI wants to move beyond rules-based automation, and to do so it has developed what it says is the industry’s first end-to-end generative AI platform for enterprise automation. The platform combines what the company calls a “multimodal large action model” with an AI agent that’s capable of using symbolic reasoning and neural network-based analysis. This foundation enables Orby to understand what’s required when faced with more complex automation requests.

The startup refers to its system as “agentic process automation” or APA, and says this is the next step up from traditional “robotic process automation” or RPA tools that dominate the industry today. With APA, Orby says it combines the high-level cognitive abilities of generative AI with the most sophisticated automations, so it can perform more complex business tasks that involve intricate planning, reasoning and adaptation.

According to Orby, when its AI agent is given a task to complete, including a set of instructions, it will autonomously generate a workflow that integrates with other, more specialized AI agents that will perform various sub-tasks that are required to fulfill the main task. These sub-tasks might include data analysis or customer interaction, the startup explained. By using various AI agents to complete each of the associated tasks, it can handle far more complex work in a range of scenarios, Orby promised.

Co-founder and Chief Executive Bella Liu said that yesterday’s RPA systems were not designed to address the most dynamic and complex enterprise workflows. “We’re blowing the lid off what AI can do for enterprises with a completely new platform that observes and learns any task performed,” she said. “It analyzes it contextually and creates automated workflows that can be immediately enacted.”

Orby says its large action models or LAMs are the secret sauce, as these are what enable its AI agents to learn, reason and act on repetitive workflow processes. As such, they can automate discovery, understanding and learning of the general patterns and relationships that exist across unstructured datasets, and this is what makes them act far more like humans, giving them the ability to handle almost any kind of repetitive task.

Enterprise Management Associates analyst Dan Twing said Orby’s LAM technology is the key to overcoming the challenges of traditional RPA, simplifying the process of workflow definition for users. “With neuro symbolic programming, its platform captures standard process flows and ensures robust exception handling to provide a resilient automation solution that now makes AI-driven automation accessible and highly efficient for the enterprise.”

The startup said the funds from today’s round will be used to bring its generative AI process automation platform to mass market and further its development.

Image: Orby AI